APPLIED MICROBIOLOGY, Sept. 1974, p. 383-391Copyright 0 1974 American Society for Microbiology

Vol. 28, No. 3Printed in U.S.A.

Effect of Dissolved Oxygen, Temperature, Initial Cell Count,and Sugar Concentration on the Viability of Saccharomyces

cerevisiae in Rapid Fermentations'TILAK W. NAGODAWITHANA, CARMINE CASTELLANO, AND KEITH H. STEINKRAUS

Cornell University, New York State Agricultural Experiment Station, Geneva, New York 14456

Received for publication 1 April 1974

By using 7 x 101 cells of Saccharomyces cerevisiae per ml with which 250 Brixhoney solutions were fermented to 9.5% (wt/vol; 12% vol/vol) ethanol in 2.5 to 3 hat 30 C, i.e., rapid fermentation, the death rate was found to be high, with only2.1% of the yeast cells surviving at the end of 3 h under anaerobic conditions. Asthe dissolved oxygen in the medium was increased from 0 to 13 to 20 to 100% inrapid fermentations at 30 C, there was a progressive increase in the percentage ofcells surviving. The ethanol production rate and total were not seriously affectedby a dissolved oxygen concentration of 13%, but fermentation was retarded by20% dissolved oxygen and still further decreased as the dissolved oxygen contentreached 100%. When the fermentation temperature was decreased to 15 C (at13% dissolved oxygen), the rate of fermentation decreased, and the fermentationtime to 9.5% ethanol (wt/vol) increased to 6 h. It was found that the higher thetemperature between 15 and 30 C, the greater the rate of death as initial cellcounts were increased from 1.1 x 107 to 7.8 x 108 cells per ml. At the lowest levelof inoculum, 1.1 x 107 cells per ml, there was actual multiplication, even at 30 C;however, the fermentation was no longer rapid. The addition of 15% sugar,initially followed after an hour by the remaining 10%, or addition of the sugar inincrements of 2.5 or 5% yielded a better survival rate of yeast cells thanwhen the fermentation was initiated with 25% sugar.

"Rapid fermentations" are defined as fer-mentations in which the ethanol content in-creases from 0 to 9.5% (wt/vol; 12% vol/vol) in 6h or less. For example, Steinkraus (11) reportedthat at a temperature of 25 C, honey solution orgrape juice (250 Brix sugar concentration) wasfermented to 9.5% (wt/vol) ethanol within a 4-hperiod. The high rate of ethanol production wasachieved primarily by using a yeast concentra-tion of 3 x 109 cells per ml. A fermentation timeof 6 h or less makes continuous fermentationfeasible. However, a problem that severely lim-its the possibility of applying rapid fermenta-tion to continuous fermentation has been thehigh death rate of yeast cells.For continuous fermentation, it is essential

that yeast viability (as a percentage of theinitial yeast charge) remains at 100% in rapidfermentation, and it is desirable that some lowlevel multiplication occurs continuously to re-place dead cells and to replace those lost duringcentrifuging of the product and during therecycling of cells to the fermentor.

l Journal paper no. 2099 of the New York State Agricul-tural Experiment Station, Geneva, N.Y. 14456.

Although there have been numerous studieson the effect of temperature on growth anddeath of cells, some having dealt with the effectof temperature on viability of yeast suspensions(14, 15), specific information regarding theeffect of temperature on viability during rapidalcoholic fermentation appears to be lacking.Atincubation temperaturesof 32 C and higher,

Saccharomyces cerevisiae, Steinberg strain 618,failed to achieve as high alcohol concentra-tions in a 250 Brix clover honey medium as itdid at lower fermentation temperatures '(11).The role of oxygen in alcoholic fermentations

has been reviewed (7). Experience has demon-strated that air or oxygen is essential for goodbrewery fermentation. Oxygen acts primarily asa terminal acceptor of electrons from the respi-ratory chain. It also acts as a yeast growthfactor. Oxygen appears to be involved in thesynthesis of oleic acid and ergosterol, whichstimulate yeast growth under anaerobic condi-tions (1).

In brewery fermentation, 20% oxygen satura-tion of the wort prior to pitching resulted in themaximal production of yeast cells (7). Above

383

NAGODAWITHANA, CASTELLANO, AND STEINKRAUS

that level, increases of cell mass were small, andbelow that level production of yeast cells wasdirectly related to oxygen supplied. It may bethat the yeast cells have the ability to retain theeffect of oxygen within the cell and use itanaerobically later (2). The amount of ethanolproduced during fermentation is independent ofthe amount of oxygen supplied prior to pitching.A high degree of aeration results in an in-

crease of biomass, i.e., multiplication andless ethanol production (8). Brettanomycesclaussenii showed a higher rate of ethanolproduction in air than in nitrogen (M. T. J.Custers, Ph.D. thesis, Delft Univ. of Technol-ogy, Delft, The Netherlands, 1940). The term''negative Pasteur effect" was introduced forthe phenomenon to emphasize the striking con-trast to the well-known Pasteur effect. A simi-lar phenomenon was reported in intact cells ofS. cerevisiae (16).

Generally, when yeast multiplication is de-sired, as with bakers' yeast, aeration is used andsugar is added in small increments. This elimi-nates repression of cell growth by high sugarconcentration and insures oxygen availability tothe cell. Yeast in a glucose medium consumedoxygen at a very slow rate until glucose wasnearly all used (13). Thus, growth in thepresence of glucose took place as though theyeast was under anaerobic conditions. Bakers'yeast ferments more than 80% of the sugar toethanol in spite of vigorous aeration (5). Thus,the classical Pasteur effect does not appear tooperate under conditions of yeast growth. Slo-nimski (9) called this the counter-Pasteur ef-fect. It describes the relationship between therate of synthesis of respirating enzymes and therate of fermentation. Yeast growing aerobicallyin glucose medium does not rapidly synthesizeits respiratory enzymes (3). Most of the energyfor growth comes from glycolysis.The counter-Pasteur effect is one aspect of

what used to be called the glucose effect and isnow described as catabolite repression (6). Thisis the inhibitory effect of glucose on the synthe-sis of enzymes, which do not immediatelycontribute to the growth of the cell. At glucoseconcentrations above 3%, cellular mechanismsfor the formation of nucleic acids and proteinsare saturated. Consequently, the cell becomesrich in adenosine 5'-triphosphate and interme-diates such as pyruvate. The cells, therefore,have no need to tap any extra energy from thecitric acid cycle, and to prevent unnecessarytapping of energy the biosynthesis of respiratoryenzymes is repressed.The level of aeration supplied to the yeast

before or during fermentation is related to the

viability of the yeast. There appears to be noinformation in the literature dealing specificallywith the effect of oxygen on the viability ofyeasts and on the rates of fermentation underconditions of rapid fermentation.Both batch and continuous fermentation data

(F. F. Pironti, Ph.D. thesis, Cornell University,Ithaca, N.Y., 1971), with an 18% (wt/vol) glu-cose concentration at 30 C, have shown that airdoes have a protective effect against productinhibition. By vacuum fermentation, it has alsobeen shown that alcohol is not the only factorcontributing to inhibition of growth or fermen-tation. Substrate inhibition has been shown tohave a relatively minor effect compared to theproduct inhibition at high concentrations ofsugar. Pironti's calculations indicated thatrates of sugar accumulation intracellularly werehigher than the corresponding rate of sugarconsumption from batch data.

This study was undertaken to determine theeffects of controlled oxygen levels, temperature,number of yeast cells inoculated, and additionof sugar in increments rather than in total onthe percentage of initial cells inoculated re-maining viable during rapid fermentation.

MATERIALS AND METHODSYeast strain. All experiments were conducted with

a beer strain of S. cerevisiae kindly supplied by MarioFrati, Genesee Brewing Co., N.Y. The yeast wasobtained as a solid paste containing from 3.85 x 109to 4.04 x 101 cells per g. The yeast was suspended incitrate-phosphate buffer (pH 6.8) and centrifuged to re-cover the cells used in the experiments.Yeast counts. Based upon the number of yeast

cells desired in the inoculated medium, a givenweight, generally 1 kg, of yeast paste was slurried with1 liter of the medium. After thorough dispersion, a1-ml sample was serially diluted and plated (twoplates per dilution) to obtain the initial viable count.The culture medium used for pour plates contained1.5% maltose, 1.5% malt extract (Difco), and 1.5%agar (Difco). The plates were incubated at 30 C for 48h, and the final colony count was taken as the averageof the two plates for the dilution containing 30 to 300colonies per plate. Calculations involving cell popula-tions were based upon viable counts. The percentageof viability was determined by dividing the viablecount at the desired fermentation time by the initialviable count.

Fermentation medium. Clover honey stored at 1 Cwas diluted with water to provide the 25° Brix sugarsubstrate. A basal level of 0.25 g of Actiferm per liter(Budde and Westermann, New York, N.Y.), a yeastvitamin mixture, was added to the fermentationmedium, and this level is referred to as X. The 0.25 gof Actiferm per liter contained: biotin, 12.5 gg;pyridoxine, 250 jig; meso-inositol, 1.87 mg; calciumpantothenate, 2.5 mg; thiamine, 5 mg; peptone, 25mg; and ammonium sulfate, 215 mg. The medium

384 APPL. MICROBIOL.

S. CEREVISIAE IN RAPID FERMENTATIONS

was also supplemented with 1.0 g of (NH4)2SO4, 0.5 gof KPO4, 0.2 g of MgCl2, 0.05 g of NaHSO4, and 5.0 gof citric acid per liter as per formula I of Steinkrausand Morse (11). The basic level of mineral salts-citricacid supplement was referred to as Y. Certain fer-mentations in this study were conducted using a 2Ylevel of mineral salt-citric acid supplement and 4Xconcentration (1 g/liter) of vitamins (Actiferm). ThepH of the fermentation medium was adjusted to 4.2,and it was pasteurized at 76.5 C for 30 min and cooledto fermentation temperature before inoculation.

Apparatus. (i) Microferm laboratory fermentormodel 214 (New Brunswick Scientific Co., NewBrunswick, N.J.) was used. A fermentor sampler(New Brunswick model S 21) was used to samplemedium during fermentation.

(ii) Dissolved oxygen (DO) controller model DO-60(New Brunswick Co.) was used in conjuction withthe above to measure and control the oxygen level.

(iii) The fermenting medium was sampled at de-sired intervals, and cell-free supernatants were ob-tained for ethanol determinations by centrifuging at14,000 x g for 10 min in a Sorvall centrifuge modelSS-3. All batch fermentations were continued untilessentially complete.

(iv) A Carle gas chromatograph, model 9000equipped with a flame-ionization detector, was em-ployed for the separation and quantitation of ethanol.A stainless-steel column (3 feet x 1/8 inch [91.44 x0.32 cm] outer diameter and 0.085 inch [0.214 cm]inner diameter) fitted into the instrument to provideon-column injection. The column packing was Poro-pak Q-S, 100/120 mesh. The column oven was oper-ated isothermally at 150 C. The carrier gas wasnitrogen at a flow rate of 20 ml/min. The combustiongases were hydrogen and air, at flow rates of 21 and300 ml/min, respectively. Sensitivity of the instru-ment for the analysis was maintained at 2 x 10- 9 Afull scale. A 2-uliter sample was injected into theinstrument with a no. 701 Hamilton syringe.The gas chromatograph was connected to a Hew-

lett-Packard integrator, model 3370A, to determinethe area of the ethanol and internal standard curves.Visual display of the chromatogram was accom-plished by means of a Houston "Omniscribe" stripchart recorder with a sensitivity of 1 mV full scale anda chart speed of 1 inch/min (2.54 cm/min).

Standard solutions of ethanol were prepared. Thestandards contained 0.5, 1.0, 1.5, and 2.0% ethanol(wt/vol). A 2% (wt/vol) acetone solution was used asthe internal standard. The areas of the peaks weredetermined by the electronic integrator. The arearatio of the ethanol to the internal standard wascalculated as follows: area ratio = area of the ethanolpeak/area of internal standard peak. The standardcurve was prepared by using the method of leastsquares. The following equation resulted in: percentethanol = a + b (peak area ratio of unknown), where a= intercept of standard curve and b = slope ofstandard curve. Knowing the intercept and the slope,a simple calculation gave the percentage of ethanol inthe unknown sample injected. When the ethanolconcentrations were high, the samples were diluted1:5 or 1:10 depending on the concentration, and 1 ml

was added to 1 ml of the internal standard. Theresulting mixture was analyzed and the percentage ofethanol obtained was multiplied by the dilutionfactor (L. R. Mattick, Agr. Exp. Station, Geneva,N.Y., and A. C. Rice, Taylor Wine Co., Hammonds-port, N.Y., unpublished data).

(v) A temperature-compensated refractometer(American Optical Co., Buffalo, N.Y.) was used todetermine the percentage of sugar and other solublesolids in degrees Brix.Method of calculating ethanol molecules pro-

duced per cell per second. The number of viable cellsat different time intervals during the course of thefermentation was determined by serial dilution andplating, as described above, and the results wereplotted.

Also, the ethanol concentrations, in grams per 100ml, were determined at the same time intervals bycentrifuging the yeast cells and analyzing the ethanolcontent of the supernatant in the gas chromatograph.The grams of ethanol produced during a specificfermentation time were changed to ethanol moleculesby multiplying the moles of ethanol produced timesAvogadro's number (6.02 x 1023 molecules per mol).The grams of ethanol produced per 100 ml was

plotted against time. By drawing a tangent at thedesired point on the curve, it was possible to get theslope or instantaneous rate of ethanol production atthat time. Dividing the rate of ethanol moleculesproduced by the number of viable cells at the sametime yielded the number of ethanol molecules beingproduced per cell per unit of time.

RESULTS AND DISCUSSIONEffect ofDO in the substrate and tempera-

ture on yeast viability during rapid fermen-tation. The percentage of DO in the mediumwas found to have a considerable influence uponthe percentage of cells remaining viable underconditions of rapid fermentation at 30 C. Thepercentage of yeast cells surviving increasedfrom 2 to 13 to 34 to 60% as the DO contentin the substrate was increased from 0 to 13 to 20to 100% (Fig. 1). The ethanol concentrationreached the desired 9.5% (wt/vol; 12% vol/vol)in the substrates containing 0 and 13% DO after3 h, but reached 8.6 and 7.3% ethanol (wt/vol)in the fermentations in which the DO wasmaintained at 20 and 100%, respectively, after5.5 h (Fig. 2).At 15 C, nearly all the yeast cells survived

rapid fermentation at 13 and 20% DO concen-tration in the substrate (Fig. 3), and, at 15 Cand 100% DO, there was actually multiplicationof the yeast with the final count reaching 195%of the initial population. The ethanol concen-tration at 0 and 13% DO content reached 9.5%(wt/vol) ethanol after 6 h at 15 C; but, similarto experiments conducted at 30 C, the ethanolcontent reached only 5.7 and 4.0% (wt/vol) after

385VOL. 28, 1974

386 NAGODAWITHANA, CASTELLANO, AND STEINKRAUS APPL. MICROBIOL.

25'BRIX 5~~~~~~~~~5'25' BRIX

3 100 ~~~~~~~~~~~~~~~3Y+x

2 2-Ea_ 402% ETHANOL W/V

a\7%ETAO WIV0% U O.~~~~~~~~~~~~~3 00 7% ETHANOL W/V

2\ 2 ~~~~~~~~~~~~~~~~~~~~~~~93% ETHANOL W/V

101%ETHANO9WIV

7j

107 I 10 'l8t l I l

0 1 2 3 4 5 0 1 2 3 4 5 6

HOURS HOURS

FIG. 1. Effect of the percentage of DO in the FIG. 3. Effect of the percentage of DO in themedium on the viability of yeast cells during rapid medium on the viability of yeast cells during rapidfermentation at 30 C. fermentation at 15 C.

6 h at 15 C in fermentations where DO wasmaintained at 20 and 100%, respectively.

Anaerobically (0% DO), only 2.1% of the cellssurvived rapid fermentation to 9.5% ethanol(wt/vol) at 30 C, whereas 84% of the cells

307C survived anaerobic fermentation at 15 C.2S°BRIX ~Whereas 13% of the clssurvived rapid fermen-

tation to 9.5% ethanol (wt/vol) at 30 C when theDO was increased to 13%, 94% of the cells

2 survived under similar conditions at 15G.Rapid fermentations carried out at 30 C with

101ow% ETHANOL WIT 101% ETHANOL wiv a cell concentration of 7 x 108 cells of S.10 96mETHANOL WIT cerevis i a eper ml anda sugar substrate of25l

3T(/v86% f HANOIBrix honey solution, wi thnutrientsadded at the

X + Y concentration with a DO content of 13%,I/ 69=% ETHANOL 73% ETHANOL reached 9.5% (wt/vol; 12% vol/vol) ethanol

DwITconcentrationin approximately 3 h (Fig. 4).s io ondi time intervals takenforsimilar

3// /S2%ETHANOL WITdfermentationsat 13% DOcarried out at 25, 20,4 ////~~~~~~~and 15 C were approximately 4, 5, and 6 h,

//v/ 1% ETHANOL W/V 101 ETHANOL arespectively. Whereas 13% of the initial cells2 - W/Vsurvive d rapidfermentationto 9.5% ethanolid/~~~~~~~~~~wt/vol) after 3 h at 30 C (13% DO), 48, 80, and0 ,W/V,_, 94% of the initial cells survived the fermenta-1H23 4 5 6 tion to 9.5% ethanol (wt/vol) at25, 20, and 15 C

HOURS (13% DO), respectively (Fig. 5). Thus, survivalof yeast cells in rapid fermentation was greatly

FIG. 2. Effect of the percentage of DO in the improved by increasing the DO content of themedium on the percentage of ethanol produced during substrate to 13% and by lowering the fermenta-rapid fermentation at 30 C. tion temperature to 15 C.

S. CEREVISIAE IN RAPID FERMENTATIONS

10

9

11

z

7

6

5

4

3

0 1 2 3 4 5 6 7

HOURS

FIG. 4. Effect of temperature of the fermentingmedium on the rate of ethanol production duringrapid fermentation at 13% DO.

These initial studies demonstrated that thelower the fermentation temperature, the betterthe retention of viability of the yeast. It also wasnoted that the greater the aeration, the betterthe retention of cell viability; and, as DOcontent approached 100% at 15 C, a good rate ofcell multiplication was reached with the yeastcount nearly doubling after a 6-h fermentation.However, as DO content increased above 13%,the rate of fermentation was inhibited, andfermentation time was extended beyond 6 h, thetime arbitrarily set as a maximum for rapidfermentation. The best conditions for rapidfermentation on the basis of the initial studieswere fermentation at 15 C and 13% DO, underwhich 94% of the cells inoculated remainedviable after the ethanol concentration reached9.5% (wt/vol) (6 h). This survival rate would notbe acceptable for continuous fermentations inwhich it is necessary to maintain 100% viabilityand, preferably, a slight rate of multiplicationto offset losses of yeast which must be recoveredfrom the output of the fermentor.

Effect of initial cell count on viabilityduring rapid fermentation. To insure thatrapid death was not due to the deficiency of anynutrients required for growth or fermentation,the medium was supplemented with twice theusual quantity of ammonium salts, phosphate,

and magnesium, along with four times the usualquantity of vitamins generally required in aregular fermentation. The DO was maintainedat 13% to insure that oxygen was not limiting.At 15 C, the cells, even at the high initial cellpopulation of 8.3 x 108 cells per ml, showedslight multiplication with the added nutrients(Fig. 6). Thus, the initial cell population byitself was not responsible for the rapid death ofcells. The rate of multiplication was progres-sively increased as the initial cell count wasdecreased from 1.61 x 108 to 5.9 x 107 to 1.03 x107 cells per ml at 15 C. However, the fermenta-tions at the two lower levels were no longerrapid. At 30 C, the higher the initial yeastcount, the greater the rate of death as the initialcell counts were increased from 1.1 x 107 to 8.0x 108 cells per ml (Fig. 7). The cells inoculatedat the high level of 7.8 x 108 cells per ml diedrapidly at 30 C, even though the nutrients andoxygen were not apparently limiting. The loss ofthe viability of cells inoculated at an intermedi-ate population (1.52 x 108 cells per ml) was notnearly as prominent at 30 C as it was in thefermentation with high cell population. Usingan inoculum in the range of 1.1 x 107 cells perml, it was found that a high level multiplicationoccurred both at 15 C and 30 C, with thehighest maximum population at 15 C (Fig. 6

z0u

-.

0 1 2 3 4 5 6 7HOURS

FIG. 5. Effect of temperature of the fermentingmedium on the percentage of cells surviving duringrapid fermentation at 13% DO.

387VOL . 28, 1974

388 NAGODAWITHANA, CASTELLANO, AND STEINKRAUS APPL. MICROBIOL.

9 and 7). Thus, the initial cell population was8 HOUR INTERVAlS found to be an important factor in determining6 whether or not multiplication, population level-5 * ing, or moderate or high death rates occurs at. F 30 C. It should be noted, however, that below3 _ 10 HOUR INTERVALS 1.5 x 108 cells per ml, the fermentation was no

E longer rapid, even at 30 C. The above resultsZ2 _ / suggested that it might be possible to get a

IHUR INTERVALS slight multiplication of yeast cells in a fermen-U / tation at 30 C by selecting an initial yeast counteo 8 r /between 1.1 x 107 cells per ml, where considera-10 0 HOUR INTERVALS ble multiplication occurred, and 1.52 x 108 cells

7

- _ y- per ml, where there was slight death at 30 C.6 t / The value of the initial cell count, which would5 > / result in a slight degree of multiplication of

13%D7O yeast cells in a fermentation at 30 C, was3-25 BRIX calculated graphically, and the cell levels calcu-

lated were confirmed experimentally.2 _ / Method of calculation. It was observed that

the percentage of viability at 30 C increasedfrom 54 to 62 to 92 to 135% (after 1 h) and 13 to

107_ ________ll___ 29 to 74 to 262% (after 4 h) as the initial cell1 2 3 4 5 6 7 counts for the fermentations were decreased10 20 30 40 50 60 70 from 7.8 x 108to5.8 x 108to 1.52 x 108tol.1 x

HOURS 107 cells per ml.FIG. 6. Effect of initial cell count of yeast cells on When the initial cell count of these fermenta-

the percentage of cells surviving during rapid fermen- tions was plotted against the percentage oftation at 13% DO and 15 C. viability at the end of 1 and 4 h, a relationship

was observed between the initial cell count and9 0 the percentage of viability at the particular8_ 30 C7 13% D.O hour. These two lines intersected at the same625' BRIX point where there should be neither death nor5'

2multiplication at an initial cell count of 7.3 x

.o 107 cells per ml (Fig. 8). The percentages of3

viability of 262%, obtained after 4 h for thefermentation initiated with 1.1 x 107 cells perE ml, was higher than the value expected from the

1- 2 10 HOUR INTERVALS

Z HOUR INTERVALS graph. However, it should be noted that this0 X <was no longer a rapid fermentation. The fer-

------H/ERVALS mentation to 7.0% (wt/vol) ethanol using 1.1 x

CIO 107 cells of inoculum per ml required 74 h, atAl 9 ~~~10 HOUR INTERVALS HOUR INTERVALS8 O9- NTERV~lwhichtime 1,536% viability (15-fold increase) of6 10 HOUR INTERVALS cells was obtained.5 _ / Fermentations were run with 6.3 x 107 and

8.2 x 107 cells per ml initial cell count to verifythe validity of the calculated value of 7.3 x 107

3 / cells per ml, which was expected to maintain astable population at 30 C. Viability figures of

2 _/ 100.8% and 99.2% at the end of 1 h and 103%and 98.8% at the end of 4 h were observed for the

7two respective fermentations (Fig. 8). This

10 confirms the calculations, which were very close1 2 3 4 5 6 7 to what was observed experimentally. However,10 20 30 40 50 60

again these were not rapid fermentations, andHOURS 58 and 60 h were required to reach an ethanol

FIG. 7. Effect of initial cell count of yeast cells on concentration of 9.2% (wt/vol) and 9.4% (wt/the percentage of cells surviving during rapid fermen- vol), at which time the viabilities were 114%tation at 13% DO and 30 C. and 90.3%, respectively. This part of the study

S. CEREVISIAE IN RAPID FERMENTATIONS

z

0

u

-aui

-."it

t

20 40 60 80 100 120 140

%VIABILITY

FIG. 8. Effect of initial cell count on the percentageof viability after 1 and 4 h offermentation at 30 C.

demonstrates that, if fermentation time is not ofprime significance, conditions of fermentationcan be selected which will enable the fermenta-tion to be conducted at 30 C while maintaininga stable yeast population. However rapid fer-mentations, i.e., a fermentation time less than 6h, could not be carried out at 30 C with mainte-nance of yeast viability.The effect of sugar concentration on yeast

viability. Adding the 25% sugar in incrementsof 2.5, 5, or even 15% initially, with 10% sugaradded after 1 h, resulted in improved viabilitiesof 83, 70, and 41%, respectively, compared with16% viability for anaerobic (0% DO) fermenta-tion started with 250 Brix sugar and continuedfor 3 h at 30 C, at which time the ethanolcontent had reached 9.5% (wt/vol; 12% vol/vol;Table 1).The use of 13% DO resulted in a considerable

improvement in viability (41 to 79%) at 30 Cwhen the sugar was added 15% at the start with10% added after 1 h (Table 1). Using an initialconcentration of 250 Brix sugar, there was no

improvement in viability with an increase ofDO from 0 to 13%. With 5% sugar increments,the viability was 70%, even in the absence ofoxygen, and it improved moderately (79%) withuse of 13% DO. Similarly, the use of 2.5% sugarincrements resulted in still higher viability(83%) anaerobically and the use of 13% DO

increased viability to 88%. The improvement inviability, with lower levels of sugar added incre-mentally, can be explained on the basis that theyeast under conditions of aeration at 30 C in ahigh-sugar medium does not rapidly synthesizeits respiratory enzymes due to catabolite repres-sion and, hence, in spite of aeration, has todepend on the glycolytic pathway for its energyrequirements. In a rapidly fermenting cell, afurther feed-back control on the glycolytic path-way may be expected due to product inhibition.This could cause a depletion of the energy levelin the system, which could ultimately bringabout a curtailing of the metabolic activitycausing death to the cell.By use of incremental feeding, the concentra-

tion of sugar in the medium has been main-tained at a low level permitting respiratoryenzymes to function whenever energy is in shortsupply. This would lower the rate of ethanolproduction in the cell and may contribute to thehigher percentage of viability maintained dur-ing incremental feeding.At 15 C, under anaerobic conditions (0% DO),

the addition of sugar in increments of 2.5, 5, or15% initially, followed by 10% after 2 h, resultedin viabilities of 95, 91, and 80%, respectively,compared with 71% viability for the fermenta-tion started with 250 Brix sugar concentrationand continued for 6 h, at which time the ethanolcontent had reached 9.5% (wt/vol). The im-proved viability, following the addition of sugarincrementally at 15 C anaerobically, also can beaccounted for by a decrease in substrate inhibi-tion. Furthernore, at 15 C a lower rate ofmetabolic activity can be expected. Thus, the

TABLE 1. Effect of adding sugar incrementally onyeast viability under anaerobic and aerobic conditions

Viability" (%)

Anaerobic AerobicIncrements" (0% DO) (13% DO)

15C, 30C, 15C, 30C,6h 3h 6h 3h

25% at zero time ......... 71 16 109 1615% at zero time, 10%

after 1 h ... 80 41 110 795% at 90-min intervals

(15 C) and 45-min in-tervals (30 C) .... .. 91 79 119 79

2.5% at 10-min intervals(15 C) and 20-min in-tervals (30 C) ....... . 95 83 122 88

aMineral salts added at level 2Y; Actiferm (vita-mins) at 4X.

b At 9.5% (wt/vol) ethanol.

389VOL. 28, 1974

NAGODAWITHANA, CASTELLANO, AND STEINKRAUS

energy requirements also may be much lessthan at 30 C, and the energy derived fromglycolysis could be sufficient to keep the cellsalive. At higher concentrations of sugar, sub-strate inhibition may affect the rate of glycoly-sis restricting the energy supply for the survivalof yeast cells. This deficiency with respect toenergy may ultimately cause death to the celleven though the death rate was lower than thatobserved at 30 C.At 15 C, under aerobic conditions (13% DO),

the addition of sugar in increments of 2.5, 5, or15% initially, followed by 10% after 2 h, resultedin viabilities of 122, 119 and 110%, respectively,compared with 109% viability for the fermenta-tion started with 250 Brix sugar concentrationand continued for 6 h, at which time the ethanolcontent reached 9.5% (wt/vol) (Table 1). Thehigher degree of cell multiplication at lowerincremental rates may again be due to theinfluence of oxygen on viability when cataboliterepression is minimal at low sugar concentra-tions.Rates of ethanol produced per cell per

second. The rates of ethanol molecules pro-duced in batch fermentations per yeast cell persecond, at different temperatures, are presentedin Table 2. Rates of ethanol production rangedfrom about 107 to nearly 3 x 10' molecules ofethanol produced per cell per s. The higher thetemperature in the range of 15 to 30 C, thefaster the rate of ethanol production. The ratesof ethanol production gradually decreased asethanol concentration increased to 9.5% (wt/vol; 12% vol/vol).An anomaly was observed in the rates of

ethanol produced per cell per second at 30 C.The rate of ethanol production per cell persecond appeared to reach its highest point at 2

TABLE 2. Ethanol molecules produced per viable cellper second in fermentations at several temperatures

at different stages of fermentation

Fermen- Ethanol molecules produced per celltation per secondtime(h) 30 Ca 25C 20C 15C

0.5 2.1 x 106 1.4 x 10' 8.8 X 107 7.3 x 10'1 2.4 x 108 1.2 x 108 7.1 x 107 5.7 x 1072 2.9 x 10' 9.2 x 107 5.2 x 107 4.2 x 1073 1.2 x 108 5.8 x 107 4.1 x 107 3.8 x 1074 1.1 x 108 4.1 x 107 3.3 x 107 3.3 x 1075 4.6 x 107 2.5 X 107 2.3 x 107 2.6 x 1076 1.4 X 107 6.4 x 10' 1.3 x 107

a Fermentation temperature.

h, although the highest rate of death wasencountered after 30 min at 30 C. We have noexplanation of this anomaly at this time.The rates of ethanol molecules produced per

cell per second have proven very useful. Forexample, at 15 C the overall average rate was4.7 x 107 ethanol molecules produced per cellper s. If the desired ethanol concentration was9.5% (wt/vol; 12% vol/vol), the yeast cell mustproduce 9.5 g of ethanol per 100 ml in the timeinterval specified. The 9.5 g of ethanol is 9.5/46mol or 1.24 x 1023 molecules. The number ofcells required to produce that many moleculesin 6 h (2.16 x 101 per s) at 15 C can becalculated from the following formula: (numberof ethanol molecules desired/100 ml)/(time inseconds x ethanol molecules produced per cellper second) = number of cells to inoculate per100 ml, i.e., (1.24 x 1023)/(2.16 x 104 x 4.7 x107) = 1.22 x 1011 cells per 100 ml or 1.22 x 10'cells per ml.

LITERATURE CITED1. Andreasen, A. A., and T. J. B. Stier. 1954. Anaerobic

nutrition of Saccharomyces cerevisiae. (II) Unsatu-rated fatty acid requirements for growth in a definedmedium. J. Cell. Comp. Physiol. 43:271-281.

2. Brown, H. T. 1914. Some studies on yeast. Ann. Bot.(London) 28(60):197-226.

3. Ephrussi, B., P. P. Slonimski, Y. Yotsoyanagi, and J.Tavlitizki. 1956. Variations physiologiques et cytolo-giques de la levure au cours du cycle de la croissanceaerobie. C. R. Trav. Lab. Carlsberg, Ser. Physiol.26:87-107.

4. Farrel, J. A., and A. H. Rose. 1967. Temperature effect onmicro-organisms. Annu. Rev. Microbiol. 21:101-120.

5. Lemoigne, M., J. P. Aubert, and J. Millet. 1954. Laproduction d'alcool et le rendement de crossance de lalevure de boulangerie cultivee en aerobiose. Ann. Inst.Pasteur (Paris) 87:427-439.

6. Magasanik, B. 1961. Catabolite repression. Cold SpringHarbor Symp. Quant. Biol. 26:246-256.

7. Markham, E. 1969. The role of oxygen in breweryfermentations. Wallerstein Lab. Commun.32(107):5-12.

8. Maxon, W. D., and M. J. Johnson. 1953. Aeration studieson propagation of baker's yeast. Ind. Eng. Chem.45:2254-2560.

9. Slonimski, P. P. 1958. Recent studies in yeast and theirsignificance in industry. Soc. Chem. Ind. Monogr.3:7-20.

10. Steinkraus, K. H. 1972. Studies on continuous productionof wine and flor sherry, p. 258. G. Terui (ed.), Proc.4th Int. Ferment. Symp. Japan, Osaka.

11. Steinkraus, K. H., and R. A. Morse. 1966. Factorsinfluencing the fermentation of honey and mead pro-duction. J. Apicult. Res. 5:17-26.

12. Stokes, J. L. 1967. Heat-sensitive enzymes and enzymesynthesis in psychrophilic microorganisms, p. 311-323. In C. L. Prosser (ed.), Molecular mechanismof temperature adaptation. American Association forthe Advancement of Science, Washington, D. C.

13. Swanson, W. H., and C. E. Clifton. 1948. Growth andassimilation in cultures of Saccharomyces cerevisiae. J.Bacteriol. 56:115-124.

390 APPL. MICROBIOL.

S. CEREVISIAE IN RAPID FERMENTATIONS

14. Van Uden, N., P. Abranches, and C. Cabeca-Silva. 1968.Temperature functions of thermal death in yeast andtheir relation to the maximum temperature for growth.Arch. Mikrobiol. 61:381-393.

15. Van Uden, N., and A. Madiera-Lopes. 1970. Concurrentexponential growth and death of cell populations of

Saccharomyces cerevisiae at super optimal tempera-ture. Z. Alug. Mikrobiol. M.P.M. 10:515-526.

16. Wiken, T. 0. 1968. Negative Pasteur effect in yeasts,p. 133-153. In A. K. Mills and H. Krebs (ed.), Aspectsof yeast metabolism. F. A. Davis Co., Philadelphia,Pa.

VOL. 28, 1974 391